nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Define factors constrains

PaoloMai94

Hi everyone!

I have a simple DoE with two continuous factors (Holes and HoleDia) and I have a constrain that is:

Holes * HoleDia^2 muste be smaller than 72000000.

Now, I wrote the following expression in the Disallowed combinations script but I still get a combination which is >72000000.

:Name(Holes) * :Name(HoleDia) ^ 2 ) >= 72000000.

Skärmbild 2025-11-03 141454.png

Accepted answers

Victor_G

Hi @PaoloMai94,

I think the constraint is not accepted as both factors are included in the constraint through the multiplication, so it may create a collinearity issue (with both factors being not independant but directly related) :

If you really want to enforce this constraint, I would instead instead recommend creating a Candidate set with all factor combinations, filtering the non-possible combinations, and then use this set to create your optimal DoE : Candidate Set Designs: Tailoring DOE Constraints to the Problem (2021-EU-30MP-78... - JMP User Community

But you may face some difficulties in the analysis if you enforce this constraint, as the factors won't be independant, so you may face multicollinearity issue with difficulties to estimate effects from each factor.

More info on creating constraints for your design : Demystifying Factor Constraints from @Jed_Campbell.

Hope this answer will help you,

All comments

mmarchandFSLR

Victor_G

Hi @PaoloMai94,

More info on creating constraints for your design : Demystifying Factor Constraints from @Jed_Campbell.

Hope this answer will help you,

PaoloMai94

Hi @Victor_G , thank you for the help. I will try as you suggested

Victor_G

Maybe there might be also another workaround to try: do you have other factors to include in the design ?
If yes, since "Holes" and "HoleDia" are closely related, you could include these parameters as one "macro" factor, for example the "Holes*HoleDia^2", with the other factors in the design ?